[racket] executable speed estimate?

From: Sam Tobin-Hochstadt (samth at ccs.neu.edu)
Date: Sun Mar 6 15:17:07 EST 2011

On Sun, Mar 6, 2011 at 3:10 PM, Charles Hixson
<charleshixsn at earthlink.net> wrote:
> Where can Racket be expected fall in comparison with other languages in
> executable speed?
>
> I'm looking at C-Ada-SBCL Lisp-Clojure-Java-Python-Ruby as a kind of a rough
> map.  I checked speeds reported by http://shootout.alioth.debian.org/, but
> no Scheme language seems to be listed.

If you look on one of comparison lists you link to, you'll find Racket:

http://shootout.alioth.debian.org/u32/which-programming-languages-are-fastest.php

And you're correct -- Racket comes in just faster than Clojure.

However, those benchmarks are highly artificial, and unlikely to be
representative of your (or anyone else's) application.  So take this
as evidence that Racket is decently fast, and that we do care about
performance, along with everything else that makes Racket a great
language to develop in.  As Matthew put it [1]

  "It's fun to run benchmarks occasionally. Now, back to working on
language design, libraries, documentation, usability..."

If you want more accurate information about how fast Racket is, try it
out on your application, or a workload that's representative.

> My rough guess is that it would fall about level with Clojure, but that's
> purely based on both being JIT compilation languages of a similar form.
> The Guide didn't seem to help, though it did give hints on how to make ones
> programs more efficient.

[1] http://blog.racket-lang.org/2010/01/benchmarks.html
-- 
sam th
samth at ccs.neu.edu



Posted on the users mailing list.